1. Field of the Invention
The present invention relates to a communication system using a packet typified by an ATM (Asynchronous Transfer Mode) communication, and particularly to a communication system capable of implementing communications based on low transmission power while maintaining a necessary error rate.
2. Description of the Related Art
A transmission error is apt to intrinsically occur in wireless communications. Therefore, a system for effecting an error correction to a payload part as well as to a header part of an ATM cell has been proposed. At this time, an error correction stronger than that for the payload part is made to the header part to lessen the abandonment of a cell due to the occurrence of an error in a header thereof. For example, a system for respectively effecting a RS (Reed-Solomon) triple error correction and a RS single error correction on a header part of each ATM cell and a payload part thereof has been described in the reference "Watanabe, Ueno, Aikawa, and Matsue `Code Error Transmission Characteristics in Wireless ATM` at Communication Society Meeting of The Institute of Electronics, Information and Communication Engineers (Japan), B-474 in 1996".
The ATM cell is a kind of packet and comprises a header part including information about a destination and attribute thereof, etc., and a payload part including the original information to be transmitted. Five bytes and forty-eight bytes are respectively assigned to them in accordance with the International Standard Specification.
When it is desired to transmit a packet through a communication channel, a frame is comprised of at least one packet and the frame is transmitted. FIG. 11 shows the structure of a common transmission frame. The transmission frame comprises a preamble 1 and data 200. The data 200 comprises a header 2, a payload 3, and a forward error correction (FEC) code for correcting a transmission error of both or one of the header 2 and the payload 3 or a cyclic redundancy check (CRC) code 4 for detecting the transmission error of both or one of the header 2 and the payload 3.
One frame is formed with at least one packet or ATM cell coupled thereto. FIG. 12 shows a conventional frame structure in which one frame is comprised of n packets. n packets 9, 10, . . . , 12 are stored in a payload 7 of the frame and a header 6 and an FEC (CRC) code 8 are added to the frame.
In a system for forming a frame for each packet and transmitting the formed frame in packet communications in which the amount of information per packet is less, the preamble 5, the header 6 and the forward error correction code or cyclic redundancy check code 8 of the transmission frame are large in overhead with respect to the payload 7 of the transmission frame and the efficiency of its transmission is poor.
In a system for compiling a plurality of packets into a transmission frame on the other hand, if the number of correction bytes increases to cope with errors produced in a frame transmission zone, then the forward error correction code increases in number and a larger number of transmission bands are required. As the transmission band increases with an increase in the overhead of the forward error correction code 8, the preamble 5 and the header 6 of the transmission frame with respect to the payload 7 of the transmission frame, transmission power required to maintain a desired error rate increases. However, no consideration has been given to a reduction in the increase in transmission power.
It is already known that a technique for coupling a plurality of packets to each other as they are to form a transmission frame, newly forming a network header and adding it to the transmission frame has been disclosed in Japanese Patent Laid-Open (Kokai) No. Hei 2-166856. This technique has a problem in that the number of bytes increases due to the further addition of new information to each packet.
It is also already known that a technique for coupling packets to each other to form a transmission frame whereas eliminating headers of packets only when the packets identical in header to each other appear, has been disclosed in Japanese Patent Laid-Open (Kokai) No. Hei 8- 32593. As compared with the aforementioned technique, this technique brings about an advantageous effect in that the number of bytes is reduced by the number of the eliminated headers. However, since the reduction in the number of bytes is limited only to the case in which the headers become the same, limitations are imposed on the advantageous effect.